karpathy

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

okkez

Prepare files and directories

mkdir -p fluentdCA/{private,newcerts,certs,crl}
touch fluentdCA/index.txt
echo 00 > fluentdCA/serial

Setup private CA

yeromin

Два Git аккаунта

В папке ~/%userprofile%/.ssh/ должен быть config, ключи (публичные *.pub и не публичные без расширения) а так же known_hosts

Config :

# Personal Github
Host github.com
User __________
IdentityFile ~/.ssh/id_rsa

SheldonWangRJT

This notes is written by Sheldon. You can find me with #iOSBySheldon in Github, Youtube, Facebook, etc.

Need

Convert .mov/.MP4 to .gif

Reason

As a developer, I feel better to upload a short video when I create the pull request to show other viewers what I did in this PR. I tried .mov format directly got after finishing recording screen using Quicktime, however, gif offers preview in most web pages, and has smaller file size.

This is not limited to developer, anyone has this need can use this method to convert the files.

pedroxs

# recursive search for keys containing "string" stripping empty results
jq '.. | objects | with_entries(select(.key | contains("ftp"))) | select(. != {})'

# same, but output propper array
jq '[ .. | objects | with_entries(select(.key | contains("ftp"))) | select(. != {}) ]'
# or
jq 'map( .. | objects | with_entries(select(.key | contains("ftp"))) | select(. != {}) )'

# transform input from {type: a, amount: 1} to {a: 1} and sum all values by type
jq '[ .[] | {(.type): .amount} ] | map(to_entries) | add | group_by(.key) | map({key: .[0].key, value: map(.value) | add}) | from_entries'

olih

Processing JSON using jq

jq is useful to slice, filter, map and transform structured json data.

Installing jq

On Mac OS

brew install jq